[Mlir-commits] [mlir] [mlir][affine] Add affine-loop-constantize-bounds transfrom pass to affine dialect (PR #211989)

[lonely eagle]

================
@@ -465,4 +465,17 @@ def AffineFoldMemRefAliasOps : Pass<"affine-fold-memref-alias-ops"> {
   let dependentDialects = ["memref::MemRefDialect"];
 }
 
+def AffineLoopConstantizeBounds : Pass<"affine-loop-constantize-bounds"> {
----------------
linuxlonelyeagle wrote:

Because affine-loop-constantize-bounds relies on the Presburger solver and has higher performance cost, whereas AffineLoopNormalize is much simpler and lightweight.

https://github.com/llvm/llvm-project/pull/211989